summaryrefslogtreecommitdiffstats
path: root/mm/memcontrol.c
diff options
context:
space:
mode:
Diffstat (limited to 'mm/memcontrol.c')
-rw-r--r--mm/memcontrol.c229
1 files changed, 218 insertions, 11 deletions
diff --git a/mm/memcontrol.c b/mm/memcontrol.c
index 6453ea5..ee97c9a 100644
--- a/mm/memcontrol.c
+++ b/mm/memcontrol.c
@@ -97,14 +97,6 @@ static const char * const mem_cgroup_stat_names[] = {
"swap",
};
-enum mem_cgroup_events_index {
- MEM_CGROUP_EVENTS_PGPGIN, /* # of pages paged in */
- MEM_CGROUP_EVENTS_PGPGOUT, /* # of pages paged out */
- MEM_CGROUP_EVENTS_PGFAULT, /* # of page-faults */
- MEM_CGROUP_EVENTS_PGMAJFAULT, /* # of major page-faults */
- MEM_CGROUP_EVENTS_NSTATS,
-};
-
static const char * const mem_cgroup_events_names[] = {
"pgpgin",
"pgpgout",
@@ -138,7 +130,7 @@ enum mem_cgroup_events_target {
struct mem_cgroup_stat_cpu {
long count[MEM_CGROUP_STAT_NSTATS];
- unsigned long events[MEM_CGROUP_EVENTS_NSTATS];
+ unsigned long events[MEMCG_NR_EVENTS];
unsigned long nr_page_events;
unsigned long targets[MEM_CGROUP_NTARGETS];
};
@@ -284,6 +276,10 @@ struct mem_cgroup {
struct page_counter memsw;
struct page_counter kmem;
+ /* Normal memory consumption range */
+ unsigned long low;
+ unsigned long high;
+
unsigned long soft_limit;
/* vmpressure notifications */
@@ -2315,6 +2311,8 @@ retry:
if (!(gfp_mask & __GFP_WAIT))
goto nomem;
+ mem_cgroup_events(mem_over_limit, MEMCG_MAX, 1);
+
nr_reclaimed = try_to_free_mem_cgroup_pages(mem_over_limit, nr_pages,
gfp_mask, may_swap);
@@ -2356,6 +2354,8 @@ retry:
if (fatal_signal_pending(current))
goto bypass;
+ mem_cgroup_events(mem_over_limit, MEMCG_OOM, 1);
+
mem_cgroup_oom(mem_over_limit, gfp_mask, get_order(nr_pages));
nomem:
if (!(gfp_mask & __GFP_NOFAIL))
@@ -2367,6 +2367,16 @@ done_restock:
css_get_many(&memcg->css, batch);
if (batch > nr_pages)
refill_stock(memcg, batch - nr_pages);
+ /*
+ * If the hierarchy is above the normal consumption range,
+ * make the charging task trim their excess contribution.
+ */
+ do {
+ if (page_counter_read(&memcg->memory) <= memcg->high)
+ continue;
+ mem_cgroup_events(memcg, MEMCG_HIGH, 1);
+ try_to_free_mem_cgroup_pages(memcg, nr_pages, gfp_mask, true);
+ } while ((memcg = parent_mem_cgroup(memcg)));
done:
return ret;
}
@@ -4276,7 +4286,7 @@ out_kfree:
return ret;
}
-static struct cftype mem_cgroup_files[] = {
+static struct cftype mem_cgroup_legacy_files[] = {
{
.name = "usage_in_bytes",
.private = MEMFILE_PRIVATE(_MEM, RES_USAGE),
@@ -4552,6 +4562,7 @@ mem_cgroup_css_alloc(struct cgroup_subsys_state *parent_css)
if (parent_css == NULL) {
root_mem_cgroup = memcg;
page_counter_init(&memcg->memory, NULL);
+ memcg->high = PAGE_COUNTER_MAX;
memcg->soft_limit = PAGE_COUNTER_MAX;
page_counter_init(&memcg->memsw, NULL);
page_counter_init(&memcg->kmem, NULL);
@@ -4597,6 +4608,7 @@ mem_cgroup_css_online(struct cgroup_subsys_state *css)
if (parent->use_hierarchy) {
page_counter_init(&memcg->memory, &parent->memory);
+ memcg->high = PAGE_COUNTER_MAX;
memcg->soft_limit = PAGE_COUNTER_MAX;
page_counter_init(&memcg->memsw, &parent->memsw);
page_counter_init(&memcg->kmem, &parent->kmem);
@@ -4607,6 +4619,7 @@ mem_cgroup_css_online(struct cgroup_subsys_state *css)
*/
} else {
page_counter_init(&memcg->memory, NULL);
+ memcg->high = PAGE_COUNTER_MAX;
memcg->soft_limit = PAGE_COUNTER_MAX;
page_counter_init(&memcg->memsw, NULL);
page_counter_init(&memcg->kmem, NULL);
@@ -4682,6 +4695,8 @@ static void mem_cgroup_css_reset(struct cgroup_subsys_state *css)
mem_cgroup_resize_limit(memcg, PAGE_COUNTER_MAX);
mem_cgroup_resize_memsw_limit(memcg, PAGE_COUNTER_MAX);
memcg_update_kmem_limit(memcg, PAGE_COUNTER_MAX);
+ memcg->low = 0;
+ memcg->high = PAGE_COUNTER_MAX;
memcg->soft_limit = PAGE_COUNTER_MAX;
}
@@ -5267,6 +5282,147 @@ static void mem_cgroup_bind(struct cgroup_subsys_state *root_css)
mem_cgroup_from_css(root_css)->use_hierarchy = true;
}
+static u64 memory_current_read(struct cgroup_subsys_state *css,
+ struct cftype *cft)
+{
+ return mem_cgroup_usage(mem_cgroup_from_css(css), false);
+}
+
+static int memory_low_show(struct seq_file *m, void *v)
+{
+ struct mem_cgroup *memcg = mem_cgroup_from_css(seq_css(m));
+ unsigned long low = ACCESS_ONCE(memcg->low);
+
+ if (low == PAGE_COUNTER_MAX)
+ seq_puts(m, "infinity\n");
+ else
+ seq_printf(m, "%llu\n", (u64)low * PAGE_SIZE);
+
+ return 0;
+}
+
+static ssize_t memory_low_write(struct kernfs_open_file *of,
+ char *buf, size_t nbytes, loff_t off)
+{
+ struct mem_cgroup *memcg = mem_cgroup_from_css(of_css(of));
+ unsigned long low;
+ int err;
+
+ buf = strstrip(buf);
+ err = page_counter_memparse(buf, "infinity", &low);
+ if (err)
+ return err;
+
+ memcg->low = low;
+
+ return nbytes;
+}
+
+static int memory_high_show(struct seq_file *m, void *v)
+{
+ struct mem_cgroup *memcg = mem_cgroup_from_css(seq_css(m));
+ unsigned long high = ACCESS_ONCE(memcg->high);
+
+ if (high == PAGE_COUNTER_MAX)
+ seq_puts(m, "infinity\n");
+ else
+ seq_printf(m, "%llu\n", (u64)high * PAGE_SIZE);
+
+ return 0;
+}
+
+static ssize_t memory_high_write(struct kernfs_open_file *of,
+ char *buf, size_t nbytes, loff_t off)
+{
+ struct mem_cgroup *memcg = mem_cgroup_from_css(of_css(of));
+ unsigned long high;
+ int err;
+
+ buf = strstrip(buf);
+ err = page_counter_memparse(buf, "infinity", &high);
+ if (err)
+ return err;
+
+ memcg->high = high;
+
+ return nbytes;
+}
+
+static int memory_max_show(struct seq_file *m, void *v)
+{
+ struct mem_cgroup *memcg = mem_cgroup_from_css(seq_css(m));
+ unsigned long max = ACCESS_ONCE(memcg->memory.limit);
+
+ if (max == PAGE_COUNTER_MAX)
+ seq_puts(m, "infinity\n");
+ else
+ seq_printf(m, "%llu\n", (u64)max * PAGE_SIZE);
+
+ return 0;
+}
+
+static ssize_t memory_max_write(struct kernfs_open_file *of,
+ char *buf, size_t nbytes, loff_t off)
+{
+ struct mem_cgroup *memcg = mem_cgroup_from_css(of_css(of));
+ unsigned long max;
+ int err;
+
+ buf = strstrip(buf);
+ err = page_counter_memparse(buf, "infinity", &max);
+ if (err)
+ return err;
+
+ err = mem_cgroup_resize_limit(memcg, max);
+ if (err)
+ return err;
+
+ return nbytes;
+}
+
+static int memory_events_show(struct seq_file *m, void *v)
+{
+ struct mem_cgroup *memcg = mem_cgroup_from_css(seq_css(m));
+
+ seq_printf(m, "low %lu\n", mem_cgroup_read_events(memcg, MEMCG_LOW));
+ seq_printf(m, "high %lu\n", mem_cgroup_read_events(memcg, MEMCG_HIGH));
+ seq_printf(m, "max %lu\n", mem_cgroup_read_events(memcg, MEMCG_MAX));
+ seq_printf(m, "oom %lu\n", mem_cgroup_read_events(memcg, MEMCG_OOM));
+
+ return 0;
+}
+
+static struct cftype memory_files[] = {
+ {
+ .name = "current",
+ .read_u64 = memory_current_read,
+ },
+ {
+ .name = "low",
+ .flags = CFTYPE_NOT_ON_ROOT,
+ .seq_show = memory_low_show,
+ .write = memory_low_write,
+ },
+ {
+ .name = "high",
+ .flags = CFTYPE_NOT_ON_ROOT,
+ .seq_show = memory_high_show,
+ .write = memory_high_write,
+ },
+ {
+ .name = "max",
+ .flags = CFTYPE_NOT_ON_ROOT,
+ .seq_show = memory_max_show,
+ .write = memory_max_write,
+ },
+ {
+ .name = "events",
+ .flags = CFTYPE_NOT_ON_ROOT,
+ .seq_show = memory_events_show,
+ },
+ { } /* terminate */
+};
+
struct cgroup_subsys memory_cgrp_subsys = {
.css_alloc = mem_cgroup_css_alloc,
.css_online = mem_cgroup_css_online,
@@ -5277,7 +5433,8 @@ struct cgroup_subsys memory_cgrp_subsys = {
.cancel_attach = mem_cgroup_cancel_attach,
.attach = mem_cgroup_move_task,
.bind = mem_cgroup_bind,
- .legacy_cftypes = mem_cgroup_files,
+ .dfl_cftypes = memory_files,
+ .legacy_cftypes = mem_cgroup_legacy_files,
.early_init = 0,
};
@@ -5312,6 +5469,56 @@ static void __init enable_swap_cgroup(void)
}
#endif
+/**
+ * mem_cgroup_events - count memory events against a cgroup
+ * @memcg: the memory cgroup
+ * @idx: the event index
+ * @nr: the number of events to account for
+ */
+void mem_cgroup_events(struct mem_cgroup *memcg,
+ enum mem_cgroup_events_index idx,
+ unsigned int nr)
+{
+ this_cpu_add(memcg->stat->events[idx], nr);
+}
+
+/**
+ * mem_cgroup_low - check if memory consumption is below the normal range
+ * @root: the highest ancestor to consider
+ * @memcg: the memory cgroup to check
+ *
+ * Returns %true if memory consumption of @memcg, and that of all
+ * configurable ancestors up to @root, is below the normal range.
+ */
+bool mem_cgroup_low(struct mem_cgroup *root, struct mem_cgroup *memcg)
+{
+ if (mem_cgroup_disabled())
+ return false;
+
+ /*
+ * The toplevel group doesn't have a configurable range, so
+ * it's never low when looked at directly, and it is not
+ * considered an ancestor when assessing the hierarchy.
+ */
+
+ if (memcg == root_mem_cgroup)
+ return false;
+
+ if (page_counter_read(&memcg->memory) > memcg->low)
+ return false;
+
+ while (memcg != root) {
+ memcg = parent_mem_cgroup(memcg);
+
+ if (memcg == root_mem_cgroup)
+ break;
+
+ if (page_counter_read(&memcg->memory) > memcg->low)
+ return false;
+ }
+ return true;
+}
+
#ifdef CONFIG_MEMCG_SWAP
/**
* mem_cgroup_swapout - transfer a memsw charge to swap
OpenPOWER on IntegriCloud